Transparency Distortion Robustness for SOTA Image Segmentation Tasks

TL;DR

This paper investigates the impact of spatially varying radial distortions on semantic image segmentation models and proposes synthetic dataset augmentation to improve robustness against such distortions.

Contribution

The study introduces a method to synthetically augment datasets with spatially varying distortions and evaluates its effectiveness in enhancing model robustness.

Findings

Distortions significantly degrade segmentation performance.

Pretraining and larger models help mitigate performance loss.

Fine-tuning on distorted images offers limited improvements.

Abstract

Semantic Image Segmentation facilitates a multitude of real-world applications ranging from autonomous driving over industrial process supervision to vision aids for human beings. These models are usually trained in a supervised fashion using example inputs. Distribution Shifts between these examples and the inputs in operation may cause erroneous segmentations. The robustness of semantic segmentation models against distribution shifts caused by differing camera or lighting setups, lens distortions, adversarial inputs and image corruptions has been topic of recent research. However, robustness against spatially varying radial distortion effects that can be caused by uneven glass structures (e.g. windows) or the chaotic refraction in heated air has not been addressed by the research community yet. We propose a method to synthetically augment existing datasets with spatially varying distortions. Our experiments show, that these distortion effects degrade the performance of state-of-the-art segmentation models. Pretraining and enlarged model capacities proof to be suitable strategies for mitigating performance degradation to some degree, while fine-tuning on distorted images only leads to marginal performance improvements.